Paper No.

326 The NACE International Annual Conference and Exposition

PERFORMANCE OF PLAIN AND GALVANIZED REINFORCING STEEL DURING THEINITIATION STAGE OF CORROSION IN CONCRETE WITH POZZOLANIC ADDITIONS

Eric I. Moreno* ancl Alberto A. Sagii6sDepartment of Civil and Environmental Engineering

University of South FloridaTampa Florida 33620

*Permanent Affiliation: CINVESTAV-M&rida, Mexico

Rodney G. PowersMaterials Office

Florida Department of TransportationGainesville, Florida

ABSTRACT

32609

The pH of the concrete pore solution is expected to be somewhat lower in concretes usingpozzolanic additions than in concrete using unblended cements. Variations in pH pore solution mayhold the key to explaining conflicting reports on the performance of galvanized rebar. To examine thatfactor, plain and galvanized rebars have been tested for over two years in concrete specimens made withcement type II, with various contents of fly ash and silica f~lme. Electrochemical impedance

measurements and sensitive polarization techniques }lave been used to measure the rate of metaldissolution in the absence of chloride contamination at two different levels of concrete moisture. Theplain steel specimens have shown little tendency for passivation in two of the cement compositions widlthe highest levels of pozzolanic addition. The galvanized steel passivated in all cases and showed

apparent corrosion current densities of less than 0.3 I-LAnear the end of the test. Implications oflength of the corrosion initiation period and on current materials selection criteria are presented.

Keywords: concrete, corrosion, fly ash, ~galvanized steel, polarizationresistance,reinforcing steel,

silica fume.

INTRODUCTION

the

Long term durability goals (75 years and more) have recently become common in constructionof reinforced concrete highway structures. This poses new challenges in controlling reinforcementcorrosion, which remains one of the most important limiting factors of the service life of concrete.

Copyright(01996byNACEInternational.Requestsforpermissiontopublishthis manuscript in any form, In part or In whole must be made In writing to NACEInternational, Conferences Division, P.O. Box 218340, Houston, Texas 77218-8340. The material presented and the views expressed in thispaper are solely those of the author(s) and are not necessarily endorsed by the Association. Printed in the U.S.A.

Long-term corrosion control approaches are relying increasingly on extending the length of the initiationperiod of corrosion (the length of time in which the reinforcement surtkce is still in the passivecondition) since the propagation stage (from the start of active corrosion until damage is externallyobserved) tends to last relatively few years. ]

Galvanized reinforcing steel bars are being newly considered for long-term durability service inchloride-induced corrosion service, specially since adverse experience with epoxy coated rebar in marinesubstructures has created interest in examining alternative corrosion control methods. There areindications that the chloride concentration threshold for corrosion initiation of galvanized steel is great :rthan that for plain steel, with consequent extension in the length of the initiation period. However,exceptional stability of the galvanized layer inside concrete is required if the coating is expected to beeffective after an initiation period that may well need to exceed 50 years. The chloride concentrationsare small during the initiation period, but t}~egalvanized layer is continuously exposed to an alkalineenvironment. Zn and Zn alloys exhibit arnphoteric behavior, corroding actively in both acidic andlikewise in highly alkaline media. In the absence of carbonation, the pH of concrete pore solutions isusually sufficiently high to prevent acidic corrosion, but excessive corrosion due to a too high pH is adistinct danger. In an extensive series of experiments Andrade and coworkers have shown that acontinuous passivating layer of calcium hydroxyzirlcate forms on the surfdce of galvanized steel whe]lthe pH is 13.3 +0.1 or below.J’4 Modern concrete formulations for high durability in se:iwater serviceinclude commonly AASHTO Type II cement (for sulfate resistance), a low water to cement ratio, aIIda high cement Fdctor with pozzokmic replacerne nt. The latter consist of fly ash (for reducedpermeability and to reduce temperature rise in mass concrete applications) and, increasingly, rnicrosilicafor early strength and reduced permeability. The poi!zolanic additions consume Ca(OH)2 and the reactionproducts may entrap alkali ions that would otherwise be present in the pore water.56 As a result, the pHof the pore water in concretes with silica fume and, fly ash additions can be significantly lower than il]comparable unblended concretes.7’8 Thus, there is a possibil ity that pozzolanic addit ions act uall y impro~Jethe performance of galvanized steel, at least by extending the initiation stage of corrosion.

The typical thickness of a galvanizing coating is in the order of 100 ~m. Therefore, corrosionaverage rates during a nominal 50-year initiation period should be well below 2 p.rn/y to ensure that a~~ycoating at all remains in place when the chloride ion concentration begins to approach the thresholdvalue. Requirements may be even more demandi [Ig since there is evidence that hot-dip galvanizi[lgcoatings provide the best corrosion protection when the outer coating layer (“q”, nearly pure Zn) is st .11in place.3’q’S)After wastage during the initial concrete curing period, the thickness of the q layer maybe only 1/5 to 1/10 of the total galvanizing coating thickness. Thus, corrosion rates during theinitiation period may need to be as small as a fraction of 1 pm/y for successful coating performance.The purpose of this investigation was to obtain an indication of whether galvanized reinforcement couldmeet these low limits during the initiation stage in concretes formulated for high durability service. Thebehavior of plain steel with surface condition representative of construction site condit ions was examintxlfor comparison.

PROCEDURE

Specimens and materials

The concrete specimens were rectangular prisms (75 mm deep, 150 mm wide, and 300 nun higl ),with two parallel rebars of the same type, protruding 50 mm out of the top end.

326/2

The cementitious material used involves were type II, fly ash class “F”, and silica fume. Allmaterials satisfied AASHTO andlor ASTM specification for Florida Department of Transportation(FDOT) construction. The mix design used are shown in Pdble I. Mix designs B, C, D, and E have atotal cementitious content representative of current FDOT concretes for marine substructure applications.

Eight concrete specimens (four with galvanized bars, and four with black bars) were cast for thisexperiment per mix design and cured for 28 days in a moist chamber. After curing, half of thespecimens were allowed to dry in the Pdbenvironment (about 60% RH, 22° C). The remaining specimenswere kept nearly water-saturated by wrapping them in plastic bags and frequently demoistening withdistilled water spray.

Two types of rebar were used, galvanized and plain steel, both size No.4 (nominal 12.5 mmdiameter). The rebar specimens were 300 mm long with a 212 mm unmasked length (surfiice area of 86cm2) directly in contact with concrete. The rebars were cast 75 mm apart in the concrete specimens.An activated titanium rod reference electrode,’() 3 mrn diameter, 50 mm long, was cast centered in eachconcrete specimen.

The galvanized rebars were hot-dip galvanized, with an avemge coating thickness of 90pm, asdetermined by metallographic examination. The thickness of the q layer was typically 25 ~m.Preparation of the bars followed generally ASTM Specification A 767, but no chromate treatment Wmused after galvanizing. To simulate construction site conditions, the plain steel bars were cast in the as-received condition, in which an orange rust layer was present. Metallogrdphic examination showed somemill scale still present beneath the rust.

pH measurements

The pH of the pore water in the concrete specimens was estimated by an in-situ leachingprocedure, described in detail elsewhere.ll A hole, 6.4 m in diameter and 25 mm in depth was drilledinto a cubic blank sample of each concrete mix, preconditioned by keeping it in a 10070 RH chambtrfor - 1 month. Distilled water (0.4 cc) was introduced in the bottom of the orifice, and the pH of thewater was periodically determined in-situ with a micro pH electrode calibmted in high pH buff~rsolutions. The pH measurements were continued over a period of several weeks until a stable terminalpH value was observed. This value was reported as the result of the test (average of two orifices).

H:ilf cell potentials

Half cell potentials were taken regularly for each rebar against the internal reference electrode.This internal electrode was periodically calibrated against an external copper-copper sulfate electrotle(CSE) placed momentarily against the external concrete surface.

Electrical resistance measurements

Concrete electrical resistance (R) was measured in each specimen between the two rebars. Thealternating current measurement was performed using a Model 400 Ni]sson soil resistivity meter. Later,R was converted to a resist ivit y value through a gec~metrical cell constant (Cc) in the following manner:

P -R”ccThe cell constant was determined to be Cc = 11 cm by calibration with rebars in a plastic cell

of similar dimensions as the concrete specimens, filled with a liquid of known resistivity.

326/3

Polarization resistance

Polarization resistance of the rebar specimens was determined by both electrochemical impedancespectroscopy (EIS), and a galvanostatic step techniq~ue.

EIS tests were conducted at an amplitude typically less than 10 mV, in the frequency interval0.001 -200 Hz using custom-made EIS instrumentation.

Galvanostatic step tests were performed simdtaneously on groups of up to 16 specimens at thesame time. The computer-controlled test technique is described in detail elsewhere. 12 The galvanostat iccurrent was selected to produce a potential deviation characteristically < 20 mV. All conversions tonominal corrosion current densities and corrosion rates were made after compensation for IR drop

effects.

RESIJLTS

Figures 1 and 2 show the evolution of concrete resistance (average of all specimens of eachconcrete type and for each environmental condition) as a function of exposure time. The specimens atroom humidity showed increasing resistance with time, as expected from continuing drying. Themoderate long-term trend of increasing resistance of the wet specimens reflects the effect of increasedcuring. Ln both environments the highest resistivit y was observed for the specimens containing thehighest pozzolanic additions, specially types D aml E (containing microsilica). The long term curi]lgof the fly-ash specimens is also evident in the increased resistivity trends for the wet environment. Theresist ivities values (p =R.1 1 cm) in the wet environment are consistent with those reported in theliterature for similar concretes. lJ

Figures 3 and 4 show the potentials (averages of quadruplicate bars) as a function of exposu retime for the plain steel and galvanized specimens in the room humidity environment. Each materialachieved a final potential that was approximately independent of the type of concrete. The potentialsreached by each material are comparable to those reported for their respective passive conditions in d“yconcrete.li The early potential history of the plain steel in concrete tnixes D and E (both with silicafume) was indicative of slow passivation, specially for mix D.

Figures 5 and G show the potential trends in the wet concretes. The galvanized specimens tended

to reach final potentials (-400 to -600 mV CSE) similar to those reported elsewhere for the passivecondition in wet concrete. 14 However, by day 750 two of the four bars in concrete mix D had reachedpotentials between -900 and -1000 mV CSE, resulting in a somewhat low average potential of -680 mVCSE. The plain steel specimens in concrete mixes A-C zind F reached potentials (- -100 mV CSE)associated with passive behavior15 early in the exposure. In contrast, the potentials in silica-fume mixesD and E were highly negative (typ. -700 mV CSE)I over the entire exposure period.

Figllres 7 and 8 show the nominal Comosion rates (see discussion below), averages l~f

quadruplicate bars for each material in the room humidity concrete tests. The galvanized rebar showed

low and continually decreasing mtes over the entire test interval, consistent with the potential trends

shown in Figure 4. By the end of the test interval, the rates in all concrete mixes were approaching thedetection limit (8.6x 10-3 win/y). The nominal corrosion mtes by the end of the test interval for the plainsteel in all corrosion mixes were also very low and approaching the detection limit (7. 1x 10--jpm/y), inagreement with the final potential values in Figure 3. The nominal rates for mixes D and E were mucl~

higher for those periods in which potentials outside the passive regime were recorded.

Figures 9 and 10 show the nominal corrosion rates for the wet concrete tests. The ga]vanizt>d

specimens showed low and continually decreasing rates, generally consistent with those reportedelsewhere for passive galvanized steellG in wet concrete. The specimens with concrete mix D (8% silicafume, 20% fly ash) showed throughout the test peric)d apparent average rates -3 times greater than thO:3ein the other concrete mixes. The plain steel specimens in mixes A-C and F showed similar, very lownominal rates over the test period, consistent with those expected from well-passivated plain steel in wetconcrete .1’ Ln contrast, the plain steel in silica-fume concrete mixes D and E showed much higharnominal rates throughout the entire test interval, coincident with the observation of very negativepotentials shown in Figure 5.

Table 2 shows the results of estimated pH measurements. The concrete mixes D and E showedthe lowest pH values, as observed also by other investigators.’,g

DISCUSS1ON

Galvanized Steel

The accuracy of the nominal corrosion rate calculated from polarization resistance measurementsis contingent upon the accuracy of the R]) measurements themselves and upon the reliability of theconversion from RP to corrosion rate. The R], calculations from experimental data are based on asimplified model of the interface that agrees generally with the result of more detailed electrochemicalimpedance tests. 12 Because of the large apparent interracial capacitances of these systems, values of Rpcan only be interpreted by extrapolation and thus should be considered as apparent values. Theconversion from apparent polarization resistance tc~nominal corrosion current density was made in thepassive regime by means of a Stearn-Geary constant B=52 mV, from whic}l metal dissol~ltion rates wereobtained assuming double ionization of Zn. The 52 mV constant was determined empirically byAndrade et al from gravimetric measurernents of passive galvanized steel using polarization resistancesobtained by comparable methods. 18 Those investigators reported accurate corrosion rate determinationswithin a factor of 2. It is therefore expected that the present nominal corrosion rates of passivegalvanized steel have a comparable level of error.

The above evaluation of corrosion rate does not invoke any specific mechanism for dissolutionof zinc in the passive state. However, some theoretical justification for the conversion approach usedmay be derived by assuming that metal loss occurs by dissolution of the passive layer at the layer-porewater interi%ce, and growth of the layer by zinc ions resulting from oxidation of the bulk lnetal at thelayer-metal interfdce. If the rate of passive metal (dissolution follows the potential-independent trel idof an ideal passive polarization curve,lg’20the electrochemical admittance of the overall anod ic react ionis zero, and the overall admittance at the polarization resistance limit is that of the cathodic reaction.Assuming it to be simple oxygen reduction, not limited by diffusion since the reaction rate is low, thepolarization resistance would be given by RI) = ~/2.3 iC,,n, where ~ is the Tafel slope of the cathodicreact ion. Therefore the Stearn-Geary constant is B = ~/2.3, which for B =52 mV yields ~ = 0.12 V,a value which is consistent with the independent 1y observed typical kinetic parameters for oxygenreduction in concrete.21

The potential and nominal corrosion rate results suggest that the galvanized steel has developedstable passivity after two years in all concrete mixes in the room humidity concrete, and in mixes A-Cand E-F in wet concrete. The potential of two of tl~e galvanized specimens in wet mix D concrete felloutside the range normally associated with passivity, but the nominal corrosion rates for those specimeluswere found to be similar to those of the other specimens in the s~itne groilp which retziined potentials in

326/5

the normal passive regime. The average nominal corrosion rates for all the galvanized specimens showa tendency to decrease with time in both environments.

Galvanized reinforcement is being evaluated in this investigation primarily for its expectedresistance to higher levels (compared with plain steel) of chloride ion contamination before initiation ofthe active corrosion. An increased contamination Ihreshold means that longer time would be requiredfor diffusion or other transport processes to build up the necessary critical chloride concentration at therebar level, thus increasing durability with respect to plain rebar construction. Therefore, the ability ofthe galvanizing coating to be present to the end of a long initiation period is essential. Keeping thelimitations of the test techniques in mind, the results in Figure 10 show nominal rates of metal wastageafter two years in the galvanized specimens in mix ID wet concrete of less than 0.3 pm/y. The nominalcorrosion rates after two years were less than about O.12 pm/y for mixes A, F, C, and E. Consicleri~lgthe observed trend for increasingly smaller nominal corrosion rates with time, the results suggest thata usable fraction of even the small q layer thickness could remain in place after 50 years of initiationservice. Penetmtion through the remaining galva~lized layers22’2swould be expected to involve longertime periods. The performance in concrete exposecl to room humidity appears to be significantly betterthan in wet concrete, with consequent increase in tile chances of survival after a long initiation period.

The results to date suggest that the pozzolanic additions tested do not create an environme[lt;Iggressive to galvanized steel. From the pH estimates in Table 2, the pozzolanes may actually assist

in keeping the pore solution away from the aggressive pH 1irnit of 13.3 suggested by Andrade et al.3There is no clear indication that too much of a reduction in pH from the pozzolan ic react ion may betaking place, but the behavior of galvanized steel in concrete mix D (low potenti:ils in some specimens)must be monitored with attention in the future.

The scope of the above discussion is limited to the survival of the galvanizing coating during the

initiation stage of corrosion, when chloride ion levels are small for a large portion of the time and

assuming that the concrete does not contain cast-in chloride contamination. The effect of rapid [y

increasin~ levels of contamination near the end of Ihe initiation stage has not been yet considered as it

is expected to involve a relatively small fraction of the total initiation time. Cases of initial

contamination and the effect of very stntill chloride levels dur-ing buildup should be examined in thefuture. The performance of galvanized steel in the propagation stage of corrosion is being investigatedin related experiments to be reported elsewhere.

Plain steel

The behavior of the plain steel bars was as expected in the case of mixes A,F and B,C, yieldingvery low nominal corrosion rates for both wet and room humidity concretes. The estimation of nominal

corrosion rates is based on empirical “B” constants (52 mV for passive steel and 26 nlV for active stee 1)reported in the literature,2q and it is expected that reasonable estimates were obtained for specimens inthe passive condition. Ln silica-fume mixtures D and E, in wet concrete, conditions normally associatedwith passivity were never reached. In dry concrete [hose conditions developed only after relatively longperiods. These anomalous behaviors are puzzling.

The nominal corrosion rates reported for the anomalous behavior in mixes D and E cannot besubstantiated at this time and must be considered only as apparent magnitudes. Severdl causes of

uncertainty exist. EIS tests of specimens in the anomalous conditions suggest the presence of a very lowapparent polarization resistance, but because of the comparatively high electrolyte resistance not enoughresolution could be obtained to provide an accurate estimate of R]) (additional testing is in progress to

326/6

determine the EIS response with greater precision). Similar inaccuracy affects the anomalous conditionR], values from the galvanostatic tests, which can only be considered as order-of-magnitude estimates.A more important uncertainty concerns the cause of the low apparent polarization resistance. It may bedue to only high corrosion rate of the steel, which would not have yet experienced passivation in thelower pH environment of the mixtures with highest and most reactive pozzolanic addition. However,even the lowest pH values observed (12.6, mix E, T~ble 2) could be expected to cause passivity aftera relatively short time. Moreover, sustained corrosion rates on the order of 25 pm/y may cause visiblecracking of the concrete cover after two years in the specimen configuration used,25 but no cracking WASobserved.

An alternative interpretation can be proposed based on the initial, red rust condition of the plainsteel specimens. If this surface layer is rich in Fe’ ‘“ ions, reduction to Fe++ may be expected as apossible cathodic reaction in concrete.2b This cathodic reaction could be matched by conversion to ahigher oxidation state of oxides in the mill scale, and/or the base metal, until an equilibrium is reachedfollowed by formation of the usual passive condition. The lower pH expected (and observed) of the highpozzolanic concrete mixes would have retarded passivation in the normal manner, allowing for thedevelopment of processes such as the one proposed. The standard Fe+++/Fe++equilibrium in aqueollssolution at pH 12-13 takes place at potentials approaching -900 mV CSE,27 which would be consistentwith the observed potential values. A red rust layer 50-100 pm thick could provide ample material tosupport electrochemical reaction rates on the order of those observed for times comparable to theduration of the test. Strain relief (and consequent lack of cracks) is conceivable given the friable andporous nature of the rust layer and the combinations of possible volume changes involved in oxifletransformations. In the wet specimens oxygen availability is limited by slow diffusion through theconcrete, and the low potential regime would be expected to last longer than in the drier specimens, :Nobserved. Because of the increasing use of silica fume for long term durability applications, resolutionof the behavior of these specimens is important. Careful monitoring for longer times and demolitionof selected specimens is pkumed to elucidate these possibilities.

CONCLUSIONS

1. Galvanized steel in chloride-free concretes with a range of pozzolanic additions reached passi~lebehavior in both wet and dry conditions. Apparent corrosion mtes after two years of exposure were ve--y

low (<0.3 pm/y) , suggesting that a significant portion of the q layer could remain in place for a periodon the order of 50 years before the arrival of a chloride contamination front.

2. The estimated pH of the pore solutions in the concretes used was below the 13.3 limit suggestedfor lack of stability of the galvanized layer. The estimated pH was lowest with the highest pozzolaniccent ent, suggesting that the pozzolanic additions might assist in improving the coating stability.However, galvanized reinforcing steel in one of the high pozzolanic mixes with silica fume showed thehighest passive corrosion rate and low potentials (but still passive corrosion rates) in a fraction of itssamples. Longer-term monitoring is continuing.

3. Plain steel controls showed passivation and very low corrosion rates zifter two years in the d-yenvironment for all concrete mixes, and in the wet conditions for mixes without silica fume. However,the plain steel bars in the wet silica fume concretes did not show potentials or apparent corrosion ratesindicative of passivity even :ifter two years. Tile presence of an initial red rust layer on the plain steelspecimens, coupled with reduced pore water pH in these concretes was proposed as a possible causeof this behavior.

326f7

ACKNOWL.EDGMENT

This investigation was supported by the Florida Department of Transportation. The opinions,findings, and conclusions expressed here are those of the authors and not necessarily those of the FloriciaDepartment of Transportation. The authors are indebted to Dr. S. Kranc for his valuable assistance inprogramming the reduction of extensive polarization data sets. One of the authors (E.1.M.) acknowledgesthe scholarship provided by the National Council clf Science and Technology (CONACYT-Mexico).

1.

2.

3.

4.

5. .

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

REFERENCES

R. Weyers, B. Prowell, 1. A1-Qadi, M. Sprinkel, M. Vorster, “Concrete Bridge Protection, Repa r,and Rehabilitation Relative to Reinforcement Corrosion,” Strategic Highway Research Program,SHRP-S- 360, March 30, 1993.

A. Sagiies, R. Powers, R. Kessler, “Corrosion Processes and Field Performance of Epoxy-CoatedReinforcing Steel in Marine Substructures,” CORROSION/94, paper no. 299 (Houston, TX:NACE, 1994).

A. Macias, C. Andrade, Brit. Corros. J., 18., 2 (1983): p. 82.

M. Blanco, C. Andrde, A. Macias, Brit. Corros. J., 19, 1 (1984): p. 41.

S. Diamond, Cem. Concr. Res., 11, 3 (1981): p. 383.

S. Diamond, J. Am. Ceram. Sot., 66, 5 (1983): p. C-82.

K. Andersson, B. Allard, M. Bengtsson, B. Magnusson, Celm. Concr. Res., 19, 3 (1989): p. 327.

U. Wiens, W. Breit, P. Schiessl, in Fly Ash, Silica Fume, Slag, and Natural Pozzolans inConcrete, ed. V.M. Malhotra, vol. 2 (Detroit, MI: ACI, 1995), p. 741.

C. Andrade, A. Macias, in Surfdce Co:itings-2, eds. A. Wilson, J. Nicholson, H. Presser (London,UK: Elsevier Applied Science, 1988): p. 137.

P. Castro, A. Sagiies, E. Moreno, L. Maldonado, J. Genesca, Corrosion, 52 (1996): in press.

A. Sagii6s, E. Moreno, C. Andrade, in preparation.

A. Sagiies, S. Kranc, E. Moreno, Electrochim. Acts (1995): in press.

N. Berke, M. Hicks, in Corrosion Forms and Control for Infrastructure, ASTM STP-1 137, ed.V. Chaker (Philadelphia, PA: ASTM, 1992): p. 207.

J. GonzAlez, C. Andrade, Brit. Corros. J., 1’7, 1 (1982): p. 21.

ACI Publication 222 R-89, “Corrosion of Metals in Concrete” (Detroit, Ml: ACI, 1989).

J. Gonzilez, A. Vtizquez, G. Jauregui, C. Andrade, Mater. Constr. (Paris), 17, 102 (1984): p.409.

326/8

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

W. Lopez, J. Gonztilez, Cem. Concr. Res., 23, 2 (1993): p. 368.

A. Macias, C. Andrade, Corros. Sci., 30, 4/5 ( 1990): p. 393.

F. Mansfelcl, in Advances in Corrosion Science and Technology, eds. M. Fontana, R. Staehle(New York, NY: Plenum Press, 1976).

S. Pyun, J. Bae, S. Park, J. Kim, H. Lee, Corros. Sci., 36, 5 (1994): p. 827.

A. Sagik%, S. Kmnc, A. A1-Mansur, S. Hierholzer, “Factors Controlling Corrosion of Steel-Reinforced Concrete Substructure in Seawater”, Nat. Tech. Inf. Service,FL/DOT/RC/0537-3523, June 1994.

S. Yeornans, Corrosion, 50, 1 (1994): p. 72.

S. Yeomans, in Corrosion and Corrosion Protection of Steel in Concrete, ed. R.N. Swamy, vol.2 (Sheffield, UK: Sheffield Academic Press, 1994): p. 1299.

C. Alonso, C. Andrade, J. GonzNez, Cem. Concr. Res. 18, 5 (1988): p. 687.

C. Andrade, M. A1OIISO,in Application of Accelerated Corrosion Tests to Service Life Predictionof Materials, ASTM STP- 1194, eds. G. Cragnolino, N. Sridhar (Philadelphia, PA: ASTM, 1994):p. 282.

J. Avila-Mendoza, J. Flores, U. Castillo, Cc}rrosion, 50, 11 (1994): p. 879

M. Pourbaix, Atlas of Electrochelnical Equilibria in Aqueous Solutions (Houston, TX: NACE,1974).

32(5/9

TABLE 1

Concrete Mix and Pro~erties

Cemlentitious Water to Measured

Mixes Cementitious Material content Cementitious Strengh

(lKg/m3) Ratio 28 day

(MPa)

A IOO%PC 333 0.55 37

F 100%PC 360 0.41 34 *

B 80% PC+20YOFA 444 0.41 46

c 70 YOPC+30YOFA 444 0.43 42

D 72’Yo PC+20YOFA+8YOMS 444 0.39 58

E 62% PC+30%FA+8%MS 444 0.39 56

PC = Portland Cement Type 11;FA= Fly Ash Class “F”; MS= Micro Silica.

* Design, measured not available.

TABLE 2

Concrete pH Pore SolutionMixes Estimated pH

A 13.02

F 13.04

B 13.00

c 13.02

D 12.92

E 12.66

326110

~ .,.,.,.,

1 “.....,.,.,.,,.,. .,...

10+::=:::::..

1

.+.,.,..,.., ,.,.,..

<

.

.—.

. ..

... .

——)4

:

;

............... .

.,..,,.,., + .

.,,,..,..,,.,., .,,

3 500 6 )

[

Mm A

tvIx F

Mix B

Mix C

MIX D

-E+

MIXE

10TIME (days)

Figure 1. Evolution of concrete resistance in the laboratory air environment, asa function of exposure time (Resistivity p - R x 11 cm).

0.1 ! 1 1 ,0 100 600 700200 300 400 500

Figure 2. Evolution concrete

TIME (days)

resistance in the wet specimens,

-9-

Mix A

*MIX F

+MIX B

*MIX C

+3-MIX D

*Mm E

1

function

of exposure time (Resistivity p - R x 11 cm).

.-;----- . ...............- &

-10001 ~ 1 i I I I ,0 100 200 300 400 500 600 700 8

TIME (days)

Figure 3. Potentials as a function of exposurelaboratory air concrete specimens.

2oo -

0-

9g -200 -Wm0m> -400-<1=zwg -600-

-800 -

-1ooo- , 1

0 100 200 :

EMIX A

Mix F

4t-

Mix B

MIX C

Mix D

Mix E

10

time for the plain steel in the

TIME (days)

[

MIx A

MIX F

Mix B

Mix C

++

Mix D

Mix E

10

Figure4. Potentials as a function of exposure time for the galvanized steel in

the laboratory air concrete specimens.

326/12

s-.E_

wco(-l

-1000: 1 1 1 I I r 10 100 200 300 400 500 600 700 8

Figure 5+ Potentials as a functionconcrete specimens.

TIME (days)

Figure 6.

EMIX A

MIX F

Mix B

Mix C

Mix D

Mix E

)0

of exposure time for the plain steel in the wet

TIME (days)

+Mix A

*Mix F

+

Mix B

*

Mix C

+

Mix D

-E?-

Mix E

10

Potentials as a function of exposure time for the galvanized steelthe wet concrete specimens.

326/13

in

[ 0.01

600 700 800

Time (clays)

Figure 7+ Evolution of the current densitiesfunctionconcreterates for

of exposure time for thespecimens. See Iext formixes D and E.

nominal corrosion mtes as aplain steel ininterpret at ion

tile laboratory airof the high apparent

0.1

Time (dlavs)

Fi~ure 8. Evolution of the current densities and nominal corrosionfunction of exposure time for the galvanized steel in the

rates as a

laboratory

air concrete specimens.

326/14

00

0

0.0001 ~~500 600 700 800

Time (days)

function ofspecimens.

Figure 9, Evolution of the current densities and nominal corrosion rates as aexposure time for the plain steel in the wet

rates as awet concrete

326/15